Film scratches diagnosis

[Ross Wilson]

Hi all, if anyone here with experience with super 16mm film damage can wade in on this I'd appreciate it.

One of five rolls came back from our lab with scratches on every frame, kind of looks like streaky rain jumping around. Intensity remains consistent. It was the last roll of the shoot so all other rolls were not affected. We used two mags and this one had been used on the same shoot for another one roll at least with no issues. 

We're shooting on an Arri SR3 Advanced, timecode mags, 24fps, 180shutter, Kodak 7222 if that helps.

The only thing of note was that at one point during the shoot, I leant a bit on the mag and the camera made a bit of a loud noise as it was running, when I immediately moved off the sound stopped. That said this issue is from frame one on this roll and the aforementioned incident didn't occur until mid roll so I don't see how the problem was caused by that. 

Anyway the lab manager is inspecting the footage but thought I'd see what you guys thought too.

 

[Doug Palmer]

Looks to me like the roll of film was loose and therefore spilling sideways , and in the darkroom someone has pushed it back straight.  Maybe some dirt or emulsion dust in there.  Or possibly this happened before the film was supplied.

[Dom Jaeger]

Yes looks like cinch marks, as Doug described. A loose roll that has some contamination between the layers which is then tightened or flattened.

[Simon Wyss]

Looks more like jump marks to me caused by the film intermittently touching something. Jump marks are (were) well known among projectionists who dealt with disrespected loop switches. Philips-Kinoton were the most prominent Sprungschrammen makers.

Since the Arriflex SR cameras use coaxial magazines through which the film runs in twist loops I have a premonition of that direction.

Edited March 24, 2022 by Simon Wyss

[Ross Wilson]

I can see how cinch marks might create a few one offs but this is every single frame multiple times a frame. Jump marks could be it, I just don't know where it would be jumping on that mag.

I've uploaded the leader section of video so if anyone wants to see it in motion they can:

https://www.dropbox.com/s/xqjhp84aj3j9cdy/FilmMarks.mov?dl=0

[Simon Wyss]

One inherent difficulty with cameras that use magazines that contain the sprocket rollers are possible slips. The loops are not stable in a manner of speaking, you can and in most cases will have a smaller lower and a larger upper loop. The transport claw rubs on a land between two perforation holes until the film falls over it, pressure from the back provided. ARRI advises to inch the mechanism in standby mode (we used to say turn slowly in older times) after attachment of a magazine to exactly achieve this. According to the instructions that would be at one frame per second as long as you hold the PHASE button pressed.

So the investigation would have to be into the loop rooms about which I know nothing, I must admit. As a camera designer I’d line those with velvet. If it’s jump scratches, you’d almost certainly find sharp edges there, perhaps burrs from manufacturing.

Cinch marks from trapped particles look a bit different, they generally don’t have that lanceolat form.

[Tyler Purcell]

I think it's probably something the film was physically touching when it was running. I don't know how, but maybe the loop size was huge and it was slamming one of the loops into the upper or lower part of the magazine throat? Something like that if there were a burr or something, would cause that type of mark. Cinching generally doesn't look like that, it usually doesn't move like the video sample shows. So I'm betting the loop was too big and it hit something in side the throat of the magazine. That's the only thing that makes sense with the footage. Of course it could have been damaged in the lab as well, but I doubt it. 

[Ross Wilson]

Thanks for all your input guys, much appreciated. 

I agree, it looks like the film is rapidly slamming against something, and I could see how a large loop would do that.

I just can't believe I made the loop that big, I've loaded so many rolls now and loop size is pretty easy to get right and check with these mags and to spot when it's too big. But I guess it's possible. Perhaps if the gears are slipping it's allowing a larger loop to form?

I've found some tests in the ASC manual to do and obviously will run some test stock through the mags. On the one hand I'm hoping the issue wont show so I don't have to get the mags serviced but then I'll never really know what it was, I think if there's no issues next time I'll have to assume it was the loop size. 

Thanks again! 

[Dom Jaeger]

Looking at the footage, the marks all slope in the same direction, but their positions are completely random. In-camera scratching tends to be repetitive, because the film is travelling a fixed path, and a loop oscillates very precisely. So I really don’t think you got this sort of random scratching from the camera. 

You would also have to really screw up the mag loading to make a loop so big and off-centre that it could hit the edge of the mag throat. And you’d also hear it.. I mean, try it out - deliberately make an oversized loop with a dummy roll and see what happens. Try to make it scratch.

20 hours ago, Simon Wyss said:

One inherent difficulty with cameras that use magazines that contain the sprocket rollers are possible slips. The loops are not stable in a manner of speaking, you can and in most cases will have a smaller lower and a larger upper loop. 

SR3 loops are exceptionally stable. You centre the loop before fitting it to the camera and at most you’ll get a one perf offset once the claw engages. Claw slippage just doesn’t happen on SR3s unless there’s a major camera fault, and if the loop size is correct you actually can’t offset it enough to hit the edges of the mag. If it did somehow slip it would jam before the film hit something. So you’d need to have a way over-size loop, and probably also have it quite offset, in order to hit anything in the mag. 

[Simon Wyss]

1 hour ago, Dom Jaeger said:

You centre the loop before fitting it to the camera and at most you’ll get a one perf offset once the claw engages. Claw slippage just doesn’t happen on SR3s unless there’s a major camera fault, and if the loop size is correct you actually can’t offset it enough to hit the edges of the mag. If it did somehow slip it would jam before the film hit something. So you’d need to have a way over-size loop, and probably also have it quite offset, in order to hit anything in the mag. 

I know. What I also know are dangers in film labs. If we assume the stock was not rewound prior to exposure nor after, our focus shifts towards the lab. As a lab worker I’d take care to prevent two things, 1) lateral slippage of parts of a roll and 2) longitudinal slippage or cinching.

Depending on the processor magazine size exposed portions are assembled into rolls of up to 2000 feet. I find that already dangerous, let alone the subject of the use of staples or tape.

These are cut marks from a sharp object the film has hit. By the angle they sit under we can say either the object stood up angled or the film loop ran twisted.

Ross, do you say the roll is scratched through its entire length?

[Tyler Purcell]

If you didn't set the film between the guides on the gate to hold it in place, the film could have fallen down quite a bit when loading. The sprocket drive is pretty locked together, one sprocket can't move independently of the other really. So I can't imagine that being the problem. 

The randomness of the scratching is due to the film not exactly lining up on the same spot every time it hit the thing that scratched it. This is very consistent with the way loops work. If they hit something hard they generally don't hit exactly in the same place, especially on 16mm. They tend to float a tiny bit and to my eyes, since there are clearly two distinct marks, it seems like the film was being scratched as it was moving down as well as the final resting position before being pulled back up by the sprocket drive. This leads me to believe it was "slammed" into something pretty hard. So it brushed something first and then slammed into something, hence the two marks that are clearly from the same object, but different places on the film. 

Now, another thought is something wrong with the guides in the sprocket drive area. Perhaps... and this is just a wild guess, but what could be happening is that the film is being pre-scratched as it's being pushed through the sprocket drive, by maybe a screw or the guide. Where I haven't seen this phenomena before, that whole assembly is super easy to access and worth examining. 

[Ross Wilson]

Just a couple of other bits of information which might help.

• Yes this reel in particular was scratched from top to tail from the very first to last frame. 
• Before this happened the same mag had been used for another reel on the same shoot and we had no problems with that one.
• Minutes after this mag ran out (the scratched reel mag) I loaded another mag and shot that with no problems at all so I'm confident it's not the camera at fault but the mag/film.
• The lab explained that they stitch all the reels together and there were five I sent in, this was the first reel to be developed yet the scratching ends immediately before the next reel starts in the scan (it's one five reel long file). So I'd be really surprised if it was the lab, they are one of the two tops labs in the UK called Cine Lab and do a lot of high end feature work.
• I've removed the throat of the mag and there's no obvious causes, clean as a whistle. 

You'd think when threading the mag with the initial few feet that it wouldn't create the exact same kind of scratches seeing as the film has been slowly pushed through the system.  So that initial mag threading section is scratched the same way the rest of the reel is. Weird. As others have said it's really hard to get the loop wrong. Even if I didn't hang the film correctly how come these scratches are on the first frames which wouldn't have travelled past the gate at all? Same for if I made the loop too big, surely the start of the film would have no scratches?

Anyway, im about to shoot a steady test with all my mags and see if this crops up again. I've also found a section in the ASC manual that explains how to feed the film through the system, inspect it, then mark the film at various points so you can see where it's starting at least. In theory.

Let you know what I find.

 

[Tyler Purcell]

3 hours ago, Ross Wilson said:

You'd think when threading the mag with the initial few feet that it wouldn't create the exact same kind of scratches seeing as the film has been slowly pushed through the system.

Maybe... but it could have taken a few seconds for the little piece of dirt that may have scratched it, to form. The act of unloading and re-loading may have dislodged it. 

[Dom Jaeger]

I really don't think this happened in-camera. 

The scratches are sloping across the frame, so they could not be caused at any point where the film travels longitudinally, which is everywhere except where the film loops through air above and below the gate. In any other area of the film travel a "piece of dirt" or other obstruction would cause scratches that are longitudinal. These are the common camera scratches you see, a straight line up and down the frame.

For the film to potentially hit the top or bottom of the mag throat it would need to be very poorly loaded, with the loop extremely offset and already touching one of the mag walls. But if you actually look at an SR mag throat, there are ridges at the top and bottom walls that are positioned so that if the film does happen to hit, it only hits on the film edges, never in the middle. 

I tried to misload this mag so that the film would scratch but I couldn't get a single mark, let alone multiple marks across the frame.

For a start, as you press the mag up to the gate, the film gets pushed back into a more centred position. I managed once to get it so badly offset that it was making a hideous noise, and still got no scratches. Even if you could somehow make it scratch, the direction would be sloping mainly longitudinally, up and down the frame rather than across like the scratches in this footage.

Again, nothing about these scratches looks in-camera to me. The fact that they are there from the start even where the film was manually loaded makes it even more obvious that this was not caused by the camera.

 

[Simon Wyss]

17 hours ago, Ross Wilson said:

The lab explained that they stitch all the reels together and there were five I sent in, this was the first reel to be developed yet the scratching ends immediately before the next reel

I have worked on processing machines, Photomec, ARRI, Debrie. What could have happened, you will perhaps not want to believe it, is that something like staple pliers or a pair of scissors were laid down in an unfortunate manner. These are often deposited in the join box (Photomec term) of the machine. The mag is put on, secured, the head of the exposed film and the foot of the running roll clipped together. Then the film is laced around rollers, pulled back taught via the mag spindle, cabinet closed, and the brake released.

Image from Cine Lab’s website

 

Here a view of a Photomec coupling cabinet, image from maintenance guide showing vacuum cleaning:

Speculation. Maybe a different machine since it’s black-and-white stock. Allen?

Do they use tape?

[Eric Eader]

Ross,

Was the film fresh, unopened newly purchased, or recanned?  That may go a long way toward explaining things.

To elaborate on what Doug Palmer stated,  there is one scenario that could cause that kind of mark from beginning to end in the rarest of circumstances and that is if the film coned out on the magazine loader and fell to the floor --- assuming a darkroom loading situation. Coning can happen in a bag as well but generally it's cleaner in the bag, (usually).

Coning can happen in the lab darkroom as well, if the lab worker will fess up.  

The center will drop out for many reasons. Hanging up on the core adaptor when adjusting for the key... loading camera; or the take-up tension is less than optimal and the core is locked to the adaptor a little more stubbornly than you would wish for while downloading the mag for lab.

To be honest, it is not likely the  entire roll would be marked, just short spaces intermittently, but that would account for perpendicular marks on the film...  stepping on it does that also, but not the whole roll.

Posting 10 seconds of film would help so we could see the timing.

You may also want to talk to Kodak.  I'm sure they have seen just about everything that could go wrong with film and may have a ready answer for you.

 

[Ross Wilson]

Thanks again guys.

Really useful information. I've posted the first 10 seconds of the roll, what you see happens throughout the reel. 

https://www.dropbox.com/s/jwd6bx0pcve3i45/Scratches.mov?dl=0

I've done scratch tests on the mags now and can see nothing resembling these marks. I'm using a loop used for looking at rock minerals so looking pretty closely. Also ran 100ft through each mag to check them, that'll be back soon.

As Dom says what is strange is that these scratches aren't vertical but horizontal. And I agree it's almost impossible to mess up a loop so bad it might do this. 

I see what you're saying Simon, if the tests come back okay it seems most likely to be not me, I sort of hope it was me really, I can fix me. Yes he did mention they tape the end of one roll to the next. I suppose between the first and second reel something could have been left in an unfortunate place like you say and then removed for the next reel. That is a nightmare I'd rather not think about.

The film was newly purchased from my usual source Eric, it came as usual and looked totally fine. I think however it would not be impossible to open and re-tape a roll. The little sticker was on the leader part that tapes it to the roll for all that's worth. It would seem like a lot of effort to run a whole roll through a camera only to re-spool it and send it back. Possible though.

I'll shoot an email over to Kodak, hopefully ill be able to rule out the mags tomorrow when the tests come back. The mystery continues... 

 

[Tyler Purcell]

Do you have a way to look at the negative under a light and see where it's at, either emulsion or back? That would help a lot diagnose it. 

I still think it has to be the camera. I still think it has to be related to the loop. 

Put simply, the loop never hits the same spot twice, hence the random pattern of it. Any other thing related to the mag wouldn't be as "randomized". Also, loop issues generally don't create vertical scratches, they generally create odd patterns. 

I can't imagine it being a lab issue, it's too consistent. Usually lab issues are inconsistent and happen in sections not over entire rolls. 

No way it's a kodak issue as the initial part of the roll is fine, so not that. 

The only other thing I can think of is debris on the guide for the counter on the magazine. Those guides are very fragile, held in place with a press fit thread system. I've fixed a bunch of them over the years and they can get a bit wonky. I'd really examine the magazine, wiggle those two guides and see. IDK how the guide roller could create the pattern because it doesn't even touch the center of the film, but maybe some debris like the tape from the head of the film, got wrapped around the guide and just grazed the film slightly? It doesn't take much to damage negative, it's super fragile. 

Anyway, those are some new ideas! 

 

[Ross Wilson]

I've just had the tests back from the lab and both mags are acting fine. No scratches or anything.

Thanks for the input Tyler. I struggle to see how it can be the loop though - when you're threading the mag you're doing it all slowly by hand and you can see nothing is touching anything. Even if the loop grew when running the camera there eIf would be a clean part at the start of the reel (the manual loading first few feet) then that might make sense. But this is on frame one to last frame

All I can think is that it must have been something between where the neg comes through from the feed to the gate or from the gate to the take up. Everything has to go through those. Frame one to the last frame. Maybe it's now been dislodged. Will definitely check those areas from now on before I load.

I'm going to talk to Kodak and Arri service to see if I can get any answers but for now im pretty stumped. 

 

[Tyler Purcell]

10 hours ago, Ross Wilson said:

Thanks for the input Tyler. I struggle to see how it can be the loop though - when you're threading the mag you're doing it all slowly by hand and you can see nothing is touching anything. Even if the loop grew when running the camera there eIf would be a clean part at the start of the reel (the manual loading first few feet) then that might make sense. But this is on frame one to last frame

The video you showed above, clearly shows the issue starting long after the roll began. Remember, the beginning of the roll is trashed from the loading process. Clearly there is no issue during that until the very second you started "running" the camera IE magazine on camera. The foot or so before that, clean as a whistle. This is why I think it has to be a camera issue and why I was thinking the loop and/or the counter roller. I'm liking my idea of the counter roller having something on it that fell off when the film was done. 

[Ross Wilson]

I can see the issue on frame one of the video.

It's faint but definitely there. If you have the video at the start and arrow key one frame at a time you can see them. I've attached a screen shot where I've upped the contrast to make it more obvious. It's obviously a lot clearer during the image though.

First frame after the hole punch at start:

 

[Doug Palmer]

It's wierd and mysterious.  Must be annoying,  with the footage you have taken,  and not knowing the cause. 

I wouldn't have thought the marks would appear linear like this if it was some dirt or something in the camera. 

[Ross Wilson]

Finally a conclusion! Might want to make a mental note of this...

I contacted the film sales company I used who put me onto Kodak. Kodak were very helpful and lunched an investigation. I just learned that it was indeed a quality control problem at Kodak which explains the fact the explanations here didn't ever seem to fit. 

Kodak have been very apologetic and from what I gather this is a rare event, they've offered to send me a few rolls and cover the cost of processing. Of course we'll have to pay out again to re-shoot because of this, but that's always a risk with film be it mechanical, lab, user or indeed supplier fault. Main thing is we know we know where the issue was.

This won't put me off shooting film but I might do a quick check on the leader of a new roll with a magnifying lens from now on. 

Hope this helps anyone who comes across scratches like this is the future!

 

[Simon Wyss]

Could have happened on a perforator. How dreadful

[Dom Jaeger]

10 hours ago, Ross Wilson said:

Finally a conclusion!

I just learned that it was indeed a quality control problem at Kodak 

Thanks for getting back to us with an answer Ross.

I thought it was pretty obvious that those scratch patterns could never have been caused in-camera, but it’s good to have an official explanation and some compensation from Kodak.